When the Yankees signed Masahiro Tanaka this winter, there wasn't any real question that they were getting an outstanding pitcher, but no one was really expecting what he's done so far. Through 22 innings, Tanaka has struck out 28 batters, good for 11.5 strikeouts per nine innings pitched (SO/9). That's some serious dominance, made all the more impressive by the fact that he never topped 10 SO/9 in any Japanese season, and posted 7.8 SO/9 just last year.
This would seem to imply that the 25-year old righty's career-high strikeout rate is unsustainable, but some have argued that Tanaka, and other Japanese pitchers, should actually expect to see an increased strikeout rate in MLB. These claims usually boil down to two points:
- Japanese pitchers use unusual deliveries and/or pitches, which MLB hitters have trouble reading. (This is the "deception" argument.)
- Japanese hitters are more focused on getting the ball in play, so MLB hitters—who are better overall, but more focused on power—are easier to strike out.
That second point in particular is pleasingly counterintuitive. Unfortunately, the numbers show that Japanese pitchers don't generally improve their strikeout rate when they enter MLB. In fact, if you remove the pitchers who were converted from the rotation to the bullpen, it looks like strikeouts tend to drop.
Baseball analysts have found that strikeout rates tend to stabilize after 70 batters faced, so we looked at every pitcher who faced at least 70 batters in Japan, and then, the next season, faced at least 70 batters in MLB. (Tanaka—who's faced 83 batters this season—already counts) That was 30 pitchers in total.* The changes in SO/9 for each of these pitchers, from their last Japanese season to their first MLB season, are shown in the chart below. A blue arrow means the pitcher's strikeout rate improved; red means it declined.
Some pitchers saw big gains in their strikeout rates (Siato, Komiyama) and others saw big drops (Okajima, Kawakami), but, as a group, the crossovers went from an average SO/9 of 7.84 in Japan to 8.01 in MLB. This represents a two percent gain, which isn't a statistically significant difference (p=0.58, two-tailed, paired).
However, seven of these pitchers—highlighted in yellow on the chart—started in Japan and relieved in America. Saito (+4.4 SO/9), Komiyama (+2.4), Ken Takahaski (+2.0), Koo (+1.4), and Yabu (+1.0) were converted to full-time relievers, and represent five of the nine largest strikeout improvements on the chart. Masahide Hasegawa and Hisanori Takahashi were converted to part-time relievers, and pitched 71 percent and 47 percent of their innings in relief, respectively. Both saw smaller strikeout gains.
As baseball fans know, relievers generally post higher strikeout rates than starters, and converting a starter into a reliever usually improves their SO/9. If we remove these seven Japanese converts, the remaining 23 pitchers went from an average of 8.27 SO/9 in Japan to 7.95 SO/9 in MLB, a four percent drop. This average is being propped up by Tanaka's hot start; the 22 unconverted pitchers before him went from 8.29 in Japan to 7.79 in MLB, a six percent drop that's close to statistically significant (p = 0.07, two-tailed, paired).
Tanaka, through three MLB starts, seems to be an aberration. This doesn't necessarily mean that his strikeout rate will drop, although I think it will. Rather, it means that if it doesn't, it has everything to do with Tanaka and his awesome splitter, and nothing to do with some intrinsic difference between American and Japanese baseball.
*That's 28 Japanese pitchers plus Wei-Yin Chen and Dae-Sung Koo, who aren't Japanese but pitched in Japan. 11 other pitchers—nine Japanese, two Korean—didn't make the "70 batters faced" cutoff: Takashi Kashiwada, Masumi Kuwate, Junichi Tazawa, and Kazuhito Tadano didn't face 70 batters in Japan the season before the joined MLB, and Kyuji Fujikama, Masanori Murakami, Tomokazu Ohka, Micheal Nakamura, Kazuo Fukumori, Chang Yon-Lim, and Sang-Hoon Lee didn't face 70 batters in their first MLB season.